Grok 4.6 (x.ai)

Article summary

Grok 4.6 is a new version of the Grok model, which builds on the previous version 4.5 with a focus on long-running agents and more ambitious interactive and visual work. It has achieved frontier intelligence across several agentic coding and knowledge work benchmarks, matching GPT-5.6 Sol on the Artificial Analysis Intelligence Index. Grok 4.6 is available in Cursor and Grok Build, with pricing starting at $2 per million input tokens and $6 per million output tokens. The model has undergone a longer supplemental training run and has been tested on projects that stretch its range and ability to sustain work over many steps.

Main themes

• Grok 4.6 release
• AI model performance
• Long-running agents
• Interactive and visual work
• Pricing and availability
• Benchmarking and comparison

What commenters say

• The release of Grok 4.6 is seen as a significant improvement over the previous version, with some considering it to be on par with or even surpassing other top models like Fable and GPT-5.6 Sol.
• Some commenters are skeptical about the timing of the release, suggesting that it may be a response to the release of Fable and other models, and that the performance gains may be due to benchmark hacking or other artificial means.
• Others argue that the performance gains are likely due to gradual improvements in the model and the training process, and that the release of Grok 4.6 is a natural progression in the development of AI models.
• There is disagreement about the value and usefulness of Grok 4.6, with some considering it to be a significant improvement and others seeing it as a downgrade from other models like Opus.
• Some commenters are concerned about the ownership and control of the model, with some expressing distrust of the company behind Grok and its intentions.
• The pricing and availability of Grok 4.6 are seen as competitive with other models, with some considering it to be a good value for the cost.